Comparative Methodology within African Linguistics

[IronFist]

The comparative method and its use

If a word X of a language A resembles in form and meaning to another word Y of a language B, three causes - and only three - can be invoked to explain this resemblance:

a) Chance: The observed resemblance is purely coincidental. We have seen that this situation is quite common and affects a fairly large fraction of the words of any two languages.

b.) Borrowing: The word of one of the languages is a loan made to the other language or to a third language C. This case is extremely frequent among languages in more or less intense contact.

c.) Genetic inheritance: the two languages inherit their respective words from the same ancestor, that is to say that the two languages derive from one and the same older language. experience shows that this case is often very difficult to distinguish from borrowing. Indeed, the two languages may have borrowed their respective words from the same language that is not the one that gave birth to them if it turns out that they derive from the same ancestral language. For example, English and German have borrowed a lot from Latin as this sample of words from these three languages shows:

Latin
cell << room >>
cuppa << cut >>
piper << pepper >>
sc (h) ola << school >>
vallum << rampart >>
vinum << wine >>

English
cell << cell >>
cup << cup >>
pepper << pepper >>
school << school >>
wall << wall >>
wine << wine >>

German
zelle << cell >>
kopf << head >>
pfeffer << pepper >>
schulle << school >>
wall << rampart >>
wein << wine >>

Similarly, modern Greek and Sango have both borrowed the Arabic words dunya << world >> and saduk << trunk >> while these three languages have absolutely no genetic connection. Borrowing between languages is therefore the rule, whether they come from a common ancestor or not. In fact each language inherits only a tiny fraction of its vocabulary directly from its predialectal ancestor. The comparative method makes it possible precisely to say whether there is genetic inheritance, borrowing or simple coincidence. For this it relies on a number of criteria, the main ones are:


1.) The regularity of phonetic correspondences: Experience shows that the phonetic changes affecting a segment of a language occur with a certain regularity. They are observed in all the words where this segment appears, unless a particular conditioning comes to affect their realization by producing other changes no less regular: in a defined context, all segement of a language evolves in a way perfectly deterministic.

It is the principle of the regularity of phonetic changes that is illustrated below perfectly by the passage from Latin to French. As can be seen in the following table, om can see that the / s / latin in initial position and immediately preceding a consonant corresponds systematically to / e / French under the same conditions. We see in this example that what counts is not the resmblance of the forms but the regularity of their correspondence: / s / and / e / are very dissimilar phonemes and yet it is the first that gave rise to the second.


Latin
stella
sc (h) ola
status
scriptura
spina

French
star
school
state
writing
spiked

Since Spanish is a language also of Latin, one should expect a similar series of correspondences between Latin and Spanish words. The table below shows that this is indeed the case.

Latin
stella
sc (h) ola
status
sciptura
spina

French
star
school
state
writing
spiked

Spanish
estrela
escuela
estado
escritura
espina

This time in / s / latin meets the segment / es / spanish, also very regularly. An obvious conclusion is now: if we only had the two modern languages of French and Spanish, we could safely conclude that the words examined here are directly inherited from a predialectal ancestor since French / corresponds regularly to Spanish / es / under the same conditions. Many other sets of correspondences - both phonological and morphological - are needed before definitively concluding the genetic relationship between these two languages. We also find that the more languages are used in the comparison, the more we are able to understand the facts and explain them. Without the intervention of Spanish, it would have been difficult to explain the appearance of / e / French instead of / s / Latin: Spanish reveals precisely the intermediate step which we actually have written records . Still based solely on French and Spanish facts, we can draw conclusions about the life and culture of the speakers of the proto-language, namely that they had the writing (see << writing >> and "school") and lived in societies organized in states.

2.) The primacy of the basic vocabulary: But the collection of correspondences to be relevant need to first relate to the lexemes least likely to be borrowed, namely the fundamental vocabulary (body components, geographical environment, phrases of reference, and many others.). Kinship, etc.).

Latin
stella
sc (h) ola
popularity
scriptura
spina

French
famous person
school
kingdom
writing
spiked

Spanish
estrella
escuela
estado
escritura
espina

English
big name
faculty
country
script
spine

 

[IronFist]

The addition of English in Table 3 exhibits that clearly everyday matches also can be discovered with borrowed lexemes. Faced with this kind of photo, the "comparative" Africanists, whether they use the technique of Jospeh Greenberg or that of Christopher Ehret, could not have hesitated for a second to include English among the Romance languages; they might even say that English is in the direction of Latin than French and Spanish are. The comparative method, consequently, rightly calls for that correspondences be first set up with lexemes from the simple vocabulary. The following table shows that this successfully makes it smooth to isolate English from the relaxation.

Table 3.1.5

French
stuffy
hand
Earth
Moon
Father

Spanish
boca
mano
tierra
luna
padre

English
mouth
hand
earth
moon
father

Professor Greenberg claims to have used such lists of words for his classifications, however he ignored the first consition, specifically the regularity of correspondence. In addition, the comparative lexemes must come from the fundamental vocabulary for every of the languages of the contrast, in any other case there is the danger of borrowing. This is exactly what can be seen with the ultimate line of Table three.1.4 where Spine English - that's indeed a borrowing from Latin - isn't always a part of the fundamental English vocabulary. This example sincerely shows that no loosening of rigor can nor ought to be tolerated, in any other case one would arrive at conclusions some distance eliminated from the reality. It is consequently essential to understand the history (as a minimum recent) and the cultures whose comparative languages are the supports a good way to:

a) better outline the extension of simple vocabulary
b.) higher establish the semantic equations among the terms of the comparison
c) better discover feasible borrowings

The comparative technique thru this description seems as a rigorous scientific technique designed to lead to indeniable conclusions that can be tested. The regularity of correspondence is in reality the result of the founding principle of all technology, the precept of causality (the identical reasons below the identical conditions produce the equal conditions produce the same effects) and ipso facto exclude chance, at the same time as the one-of-a-kind use simple vocabulary makes it possible to exclude borrowing from the opposite two possible reasons. It is therefore the simultaneous use of collection of ordinary correspondences and their limit to manifestly inherited lexemes (fundamental vocabulary) which makes it viable to demonstrate genetic inheritance. The reconstruction of proto-phonemes can then take place, but here too severe rigor is required due to the fact we need to enchantment to the precept of economic system which will achieve a phonological machine as simple as viable and capable of explain all statistics.

Kinship is absolutely tested only with the status quo of morphological correspondences similarly to phonetic correspondences. But right here again we need series of correspondences of morphemes - and no longer phonemes - respecting however the regulations of correspondences set up in the preceding stage. In this field both, one is not secure from borrowing or even mere coincidences. Again English will serve as an instance. The English suffix of verb -ate derivation (create, terminate, etc.) comes directly from the Latin -atus participle of borrowed verbs (creare, Terminare) and has become completely included with the English morphology since it permits to create new verbs (orchestrate, assassinate, etc); the equal is genuine of the suffix adverbial -in a position - borrowed from French - and which applies even on simply Germanic words (see eatable, washer-friendly, and so forth.)


The lifestyles of terms together with "African linguistics" or "Africanist" reflects the notion in a certain harmony of African languages. In truth, from the point of view of typology, these have many not unusual traits. In the e book with the aid of B. Heine & D. Nurse (2001) exquisite articles have been devoted to this challenge, considerably that of G. Clements on phonology (pp. 123-160), that of D. Dimmendaal on morphology (pp. 161-193) or that of D. Creissels on typology (231-258).

But what topics maximum to the comparatist is the extreme variety of African languages and their a couple of interactions among them. We have seen within the preceding chapters that the special classifications proposed up to now don't have any historic significance and can not be of a great deal assist, as African languages are globally similar to every different and at the equal time show sturdy neighborhood convergences. Due to touch. In addition, almost all of these languages are only recognized very recently and simply nothing is thought of their past histories. What languages have they been in contact with, wherein and at what time, for a way lengthy, in what forms?

 

[IronFist]

The linguistic state of affairs of the continent is therefore one of the most complex because contacts ought to were common among all African languages and at all times. How else can one give an explanation for that capabilities inclusive of tones or ATR (Advanced Tongue Root) vocal concord have been discovered at some stage in the continent?

In these conditions and given what we recognize approximately the phenomena of contact, it is not less rigorous than it's far essential for the historical have a look at of the African languages as Greenberg and his disciples have ceased to cry, c is exactly the other. Yes, it takes more rigor than it took to show the kinship of the Indo-European languages between them. The application of the comparative approach will have to be observed by using new restrictive rules designed to make certain that the kinships to be installed are of a genetic nature. In this undertaking we are able to take gain of the multiplicity of African languages since we can have at our disposal a large number of linguistic records to investigate. This linguistic diversity will fill a little the shortage of written documents, due to the fact every word inherited from the predialectal ancestor has its personal history in every of the girls' languages and evolves at its personal speed. Moreover, orality, that's an vital feature of African languages and culture - along with ancient Egypt, which placed incredible emphasis at the word and the pressure of words, imposed constraints on the company of words and speech to facilitate their memorization and transmission. This linguistic evolution extra than might the overall use of writing. We will see that certainly many of the maximum historic Egyptian archaic.

 

[IronFist]

Conditions of application of the comparative method in Africa

Before defining the protocol of application of the comparative method to African languages, let us summarize all that we know in a sure way about these languages both in terms of their structure and their reparations in time and space, without forgetting the difficulties they present. These certainties are few. Here they are :

a.) There is at present no general (genetic) classification of African languages.
b.) A careful examination of the documents currently available shows that, apart from Semitic, Berber and Khoisan, all the African languages - which we will now call African-Negroes in the remainder of this book - generally offer the same structure. phonological and morphological and the same evolutionary dynamic (gradual disappearance of grammatical gender, dueling, suffixal conjugation, etc.)

c) There is a prehistoric language - which we will call Negro-Negro as a result of Th. Obenga - who has historically fathered egyptian (ancient and Coptic), Wolof (CA Diop: 1989) and Bantu languages (Th. Obenga)

d) There are many concordant clues indicating that otherwise at least most of the other current Negro-African languages are also derived from this dialect ancestor.

e.) The demonstration of kinship and the reconstruction of Black Negro are difficult to carry out because of the high number of languages involved, their mutual influences and the age of the period of unity.

 

[CEITEDMOFO]

subs to this

 

[IronFist]

subs to this

respect

 

[Neuromancer]

subs

 

[IronFist]

more from my notes.

Exclusive use of only facts actually attested

To be as rigorous as possible, it is essential to avoid using proto-language as terms of comparison. Only the facts (phonological, grammatical and lexicological) attested must be taken into account. The reason is that reconstructions are never final and can be challenged overnight. F. de Saussure remarks in this connection in his Course that "Indo-European has been successively assumed * akvas, * ak1vas, * ek1vos, finally * ek1wos; only one remained unchallenged, as well as the number of phonemes. >> (Saussure: 1995, p 301). In Africa itself, there are at least two reconstructions of the proto-bantu (Guthrie's, which is the author's opinion, the simple translation of the phonetic correspondences, and that of Meeussen, who claims to render the exact forms of the proto-language). In reality, no serious comparatist uses reconstructed forms as the starting point of a comparative approach. Only "comparative" Africanists use and abuse them in their mega-classifications, going so far as to compare reconstructions of reconstructions with other reconstructions or duly attested terms. Indo-Europeanists have never done so. It is remarkable that Obenga who has proto-bantu, however, has not been tempted to use it in his methodology, preferring to rightly compare a single Bantu language with the two states of the Egyptian (ancient Egyptian and Coptic) as requires the comparative method

 

[IronFist]

^^

 

[IronFist]

exclusion of any element whose etymology can not be established from the language to which it belongs

This principle is at the heart of the present method and its strict application makes it possible to exclude any loan without dispute. Although extremely restrictive, it is absolutely necessary in the African linguistic context where linguistic contacts are spread over a period of several tens of thousands of years.

On the lexicological level this principle is well known to the comparatists since it is he who makes it possible to know if the resemblance between two lexemes of two different languages is purely fortuitous or not. If the etymologies of the two lexemes refer to words that have nothing to do with each other, it is because it is a fortuitous resemblance. It is now necessary to extend this highly successful etymological technique to morphology by taking advantage of recent grammaticalization gains. For when two morphemes have the same form and function in two different languages, the comparatists usually hasten to deceive that they are genetically related. However, this formal and functional resemblance can be, as for lexemes, due to chance, borrowing or even genetic inheritance. Indeed grammatical morphemes are often borrowed. S. Thomason in his latest book Language contact: an introduction (2001) clearly shows that nothing in the language is immune to borrowing: "What can be adopted by one language from another? The short answer is anything. various claims may be found in the literature to the effect that they are unobtrusive, but counterexamples can be found to have been made to date "" Even entire paradigms ( as those of personal pronouns) are borrowed and have actually been borrowed. Therefore, one must be extremely careful when drawing genetic conclusions about similarities of such paradigms, especially when it comes to personal pronouns. In this connection we can cite the resemblance - purely fortuitous, of course - between the singular personal pronouns of Welsh and the ancient Egyptian: Welsh i "I" / Ancient Egyptian i id. ; Welsh ti "you" / Egyptian ancient k "you (masc.)", t "you (fem.)"; Welsh ef "he" / egyptian ancient f "he". it is clear that faced with such a correlation, Joseph Greenberg and Hamito-Semitizantes would not have hesitated a second to classify Welsh in the mythical Afro-Asian phylum if this language was spoken not in Europe but in North Africa. Moreover, it can be seen that personal pronouns, far from being stable as is often asserted, are very versatile. Modern Indo-European languages confirm this quite well

 

[IronFist]

More from my notes:

The number one position of this principle in African comparative research is obvious. Let us not forget, for example, the formation of the causative in some of African languages: Egyptian, Semitic and Berber use a s-prefix, while Cushyticus makes use of a suffix to the chamito-semiticants to advise a form. Proto-Hamito-Semitic. But as on the identical time in Bantu a causative suffix of the equal form because the Cushytic, Obenga reconstructs for Negro-Egyptian a prefix * s- of the identical shape as the prefix of proto-Afro-Asiatic. Who says the truth? There is virtually nothing to permit the departure between these positions. The best way to give an indisputable solution is to make in every of the languages (and not families) the etymology - that is, the history - of this coveted affix. If it may be explained with the aid of the resources for each of the languages - which might be imagined to be continuations of the equal proto-language - then we will communicate of inheritance. In the opposite case it's miles borrowing or fortuitous convergence, even though the possibility of inheritance cannot be truly rejected. All a good way to be tested inside the second part of this ebook way to this rigorous approach is completely right Obenga and condemns once more Afro-Asian, Semitic and Berber failing this test.

It is consequently an exceedingly fertile method, particularly nicely adapted to African languages but which may be applied to all languages. It is the rigorous observe of grammaticalization methods that makes it possible to arrive at these outcomes, and many comparatists have greater or less implicitly used it. Consider for instance the formation of the feminine in two given languages and which is carried out through affixes. If the two affixes have similar bureaucracy - as is the case between Berber and Semitic - this does not show that they're inherited from a not unusual ancestral language, due to the fact it could be borrowed from one of the languages or a third language, especially if each languages had been once in contact. On the other hand, if we discover in one (or both) languages a lexeme of the identical shape which means "woman", "female" or some thing that refers to this semantic discipline, then we can say that this affix is inherited if similarly Phonetic correspondences with this lexeme are everyday.

It is consequently the ideal technique for distinguishing languages which have started to use a grammatical morpheme within the middle of the day from those who have without delay inherited it from their predialectal ancestor. French has for this reason coined the morpheme of negaton no longer [pa] from the root of the words which means "passage" (not, bypass, etc.). We can not therefore say that this morpheme is genetically related to the negation morpheme of the sango pe, which is quite a special story.

 

[KingsOfKings]

going to read up on this later.

 

[KingsOfKings]

good thread btw.

 

[IronFist]

thanks

 

[IronFist]

continued

It's far consequently an incredibly fertile approach, mainly properly tailored to African languages however which may be applied to all languages. It is the rigorous have a look at of the methods of grammaticalization which makes it feasible to arrive at those results, and plenty of comparatists have more or less implicitly used it. Consider as an example the formation of the feminine in two given languages and that's done through affixes. If the 2 affixes have similar forms- as is the case among Berber and Semitic - this does not prove that they're inherited from a commonplace ancestral language, due to the fact it can be borrowed from a 3rd language, in particular if two languages were once in contact. On the other hand if we find in a single or both languages a lexeme of the identical significant "lady", "woman" or some thing that refers to this semantic discipline, then we are able to say that this affix is inherited if in addition the phonetic correspondences with this lexeme are regular.